基于Webots汽车起重机工作场景的感知与路径规划研究

发布时间：2018-10-10 17:05

【摘要】：汽车起重机的吊装过程繁琐复杂,吊装方案制定也十分困难,而吊装方案的好坏直接影响到吊装的安全及成本,而汽车起重机利用自身的优势,在吊装工作中占了很大比重。面对日益复杂化的吊装物与吊装环境,单纯依靠人工经验完成吊装工作,让吊装的安全性及氋效性都会降低。为解决吊装方案制定繁琐,提高吊装安全等问题,基于虚拟吊装仿真技术应运而生。本文从起重机吊装安全的角度入手,基于Webots仿真软件,研究汽车起重机三维虚拟吊装环境下的工作场景感知和路径规划算法,了解吊装物、障碍物、起重机三者的关系,在障碍物间寻找一条最优路径,保证吊装安全。主要的工作包括:逡逑首先,详细分析了汽车起重机实际吊装工艺、主要技术参数、基本操作、感知系统和吊装路径参数等,通过对吊装路径影响因素的分析,在Webots中建立吊装环境的模型,通过起重机的感知系统,获取吊装场景图形,对吊装工作场景信息进行感知,为路径规划研究做准备。逡逑其次,根据起重机吊装运动特点及站位来确定吊装起始点和目标点。提出基于改进遗传算法的路径规划算法,通过作业场景模型和汽车起重机位姿空间模型,将吊装路径空间分割为多个路径平面进行处理,建立路径平面模型,最后利用遗传算法原理建立的路径点信息模型确定起重机吊装路径。该算法通过对场景模型添加包围盒属性(即边界对象属性)来提高搜索精度,用记忆算子提高计算速率和收敛速度。最后,在虚拟吊装软件平台上,采用VRML97编程语言,完成虚拟吊装的仿真测试。
[Abstract]:The lifting process of automobile crane is complicated and it is very difficult to make the lifting scheme. The quality of the hoisting scheme directly affects the safety and cost of the hoisting, and the truck crane takes a large proportion in the hoisting work by using its own advantages. In the face of the increasingly complicated hoisting and hoisting environment, only relying on artificial experience to complete the hoisting work, so that the safety and efficiency of hoisting will be reduced. In order to solve the problem of making the hoisting scheme tedious and improving the safety of hoisting, virtual hoisting simulation technology emerges as the times require. From the view of crane hoisting safety, based on Webots simulation software, this paper studies the working scene perception and path planning algorithm of automobile crane in 3D virtual hoisting environment, to understand the relationship among hoisting objects, obstacles and cranes. Search for an optimal path between obstacles to ensure hoisting safety. The main work includes: first of all, the actual hoisting technology, main technical parameters, basic operation, sensing system and hoisting path parameters of truck crane are analyzed in detail. The model of hoisting environment is established in Webots. Through the perception system of crane, the lifting scene graph is obtained, and the information of hoisting work scene is perceived, so as to prepare for the research of path planning. Secondly, the starting point and the target point of hoisting are determined according to the movement characteristics and station position of crane hoisting. A path planning algorithm based on improved genetic algorithm is proposed. The lifting path space is divided into multiple path planes and the path plane model is established by using the job scene model and the position and pose space model of truck crane. Finally, the path information model based on genetic algorithm is used to determine the crane hoisting path. By adding bounding box attributes (i.e. boundary object attributes) to the scene model, the algorithm improves the search accuracy and improves the computing rate and convergence speed by using memory operators. Finally, on the platform of virtual hoisting software, VRML97 programming language is used to complete the simulation test of virtual hoisting.
【学位授予单位】：北方民族大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TH21;TP391.9

【参考文献】